The ANOVA on the data from the 1000–2000 msec interval gave rise to a significant interaction between discrimination difficulty, subsequent
memory and scalp location [F(1, 27) = 6.82, p = .015], which was further modulated by electrode site [F(5.2, 140.4) = 3.03, p = .011]. Separate analyses in each discrimination difficulty condition revealed an interaction between subsequent memory and scalp location for the easy condition [F(1, 27) = 11.73, p = .002]. This interaction reflected a negative-going subsequent memory effect at anterior [F(1, 27) = 5.32, p = .029] but not posterior (p = .482) locations. Visual and auditory cues involving a difficult discrimination www.selleckchem.com/products/SP600125.html did again not elicit significant encoding-related effects (p > .216). No significant effects emerged in proximity of word onset for either difficulty condition (p > .116). As typically observed (Friedman and Johnson, 2000), words that were later remembered elicited more positive-going PD-0332991 cell line waveforms over frontal scalp sites than words that were later forgotten (Fig. 5). Encoding-related activity elicited by words was quantified by measuring mean amplitudes in the 700–1200 and 1200–1900 msec intervals. These intervals were similar to those used to quantify post-stimulus subsequent
memory effects in previous investigations (e.g., Galli et al., 2011; Otten et al., 2006, 2010) and captured the effects in the group averaged waveforms for all relevant conditions. The ANOVA revealed a significant interaction between subsequent memory and scalp location in both latency intervals [respectively F(1, 27) = 7.04
and 9.13, p = .013 and .005]. Subsequent memory effects were largest over anterior scalp sites, but significant at both anterior locations [F(1, 27) = 16.83 and 18.91 for the two intervals, both p < .001] and posterior locations [F(1, 27) = 10.49 and 8.13, respectively, OSBPL9 p = .003 and .008]. No interactions involving modality or difficulty emerged (p > .117). The findings indicate that encoding-related activity before an event is sensitive to the degree to which processing resources are available. Electrical brain activity elicited by a cue presented just before word onset predicted later recall of the word, but only in a low demand situation when a concurrent task was easy to perform. Participants were asked to memorize short lists of words while making perceptual discriminations on cues that preceded the words. Discrimination difficulty was manipulated across lists by making the cues more or less similar to one another. The performance data show that cue discriminations were indeed faster and more accurate in the easy condition. The lower demands in this condition may have left sufficient opportunity to also engage brain activity that affects the encoding of the upcoming word. Accordingly, activity before word onset predicted later memory of the word.